Near a hearth at Pincevent, south-east of Paris, people sat down to make stone tools about 14,000 years ago. The flint they left behind preserves a small social scene. Some hands struck long, controlled blades from carefully prepared cores. Others made the short, clumsy removals of beginners. At several Palaeolithic sites, archaeologists have found similar patterns: expert knappers working beside novices, waste arranged in ways that suggest watching, practice and perhaps instruction. The scene reveals something older than job titles. People of different abilities were contributing different things, and knowledge itself was being divided, carried and passed on.
The hunter and the gatherer are only the most familiar figures in this story. A foraging camp also had children learning to dig, track and make; adults at different peaks of strength and knowledge; people processing hides, maintaining fire, shaping tools, finding water, tending injuries and caring for infants who could not care for themselves. The work changed with season, landscape, pregnancy, injury, success and need. Some tasks were done by nearly everyone. Some fell repeatedly to the people who were best placed or most practised to do them.
This was a division of labour at the level of tasks and roles, long before a world of occupations. The first division was a flexible arrangement inside a group whose members depended on one another. Work was divided by age, skill and circumstance; often, though never perfectly, through gendered roles partly shaped by reproductive biology; sometimes by reputation or preference. The arrangement worked because results moved between people through provisioning, sharing and care.
Here, sex refers to biological characteristics and reproductive constraints; gender refers to the social roles and expectations organised around them. The evidence rarely lets us see either cleanly in deep time. Archaeology may estimate biological sex from human remains, but it cannot simply recover a person's gender or the meaning their community gave a task. The terms are related in this history, but they are not synonyms.
The distinction matters. If we reduce the beginning of work to “men hunted and women gathered”, we miss most of the economy and mistake one recurring pattern for the whole invention. The deeper question is larger: how did a species of individually vulnerable generalists learn to survive through complementary skill?
Evidence horizons, not an origin date
No surviving object tells us when labour was first divided. These points show when particular kinds of evidence become visible and when one influential interpretation entered the modern story.
More than 2.5 million years ago — Tools without job descriptions
Deep time. Oldowan tools at Gona establish skilled production, but not who made them, how tasks were allocated or whether roles persisted.
c. 50,000–12,000 years ago — Broader toolkits and wider networks
Upper Palaeolithic. Regional evidence includes more diverse technologies and long-distance material movement. Kuhn and Stiner connect some of this change to differentiated roles; the hypothesis remains debated.
c. 14,000 years ago — Novices and experts at Pincevent
Learning. Differences in stone-working skill preserve a plausible scene of practice beside expertise, not proof of formal teaching. The wider evidence is reviewed by Milks and colleagues.
c. 9,000 years ago — A female burial with hunting equipment
Burial evidence. The Wilamaya Patjxa burial challenges a universal male-hunter assumption, while burial goods still require cautious interpretation.
1966–1968 — “Man the Hunter” becomes an origin story
History of interpretation. A Chicago symposium and its published proceedings helped elevate male-coded hunting into a simplified explanation of human evolution.
The dates are approximate and regionally uneven. The first four entries concern surviving evidence; the final entry concerns how that evidence was interpreted. None marks the first division of labour.
Before jobs, there were tasks, roles and practices
“Division of labour” can mean several different things, and later chapters will depend on keeping them apart.
At the simplest level there is task differentiation. One person gathers wood in the morning, repairs a net in the afternoon and watches a child while another cooks. Work is divided in time, but people remain generalists. There is also role specialisation: particular people repeatedly hunt, heal, navigate, organise a drive or make the finest blades because they have the opportunity, body, knowledge or inclination. Finally there is occupational specialisation, in which a recognised role absorbs much of a person's working time and its product is exchanged for the food and work of others. The first two are ancient. The third emerged gradually as storage, dense settlement and institutions made it easier to sustain.
This first chapter is about the older forms. In a subsistence economy, work flowed through the day and year: finding and processing food, making and repairing what the group used, moving camp, caring for the young, maintaining relationships and knowledge. It extended beyond a clock-defined working day and narrow ideas of survival to include play, teaching, ritual, art and social obligation. Modern economies separate these categories more sharply than many people may have experienced them.
We know this past through an uneven archive. Stone, bone and teeth survive; baskets, cord, wooden digging sticks, plant foods, stories and most evidence of care have disappeared. An archaeologist can often see that a blade was made skilfully while its social destination remains obscure. A skeleton may record repeated strain, but seldom the name a community gave the task that caused it. Ethnographic studies of recent foragers supply richer detail from communities with histories of trade, displacement, colonisation and adaptation of their own.
The evidence supports a history of accumulating capacities and arrangements. Oldowan stone tools show that hominins were making things more than 2.5 million years ago, although they leave who made what for whom unresolved. Steven Kuhn and Mary Stiner have argued that the broader diets and technologies visible in the early Upper Palaeolithic reflect more differentiated roles among Homo sapiens than among Neanderthals, perhaps giving modern-human populations a demographic advantage. Their proposal is important and disputed. The archaeological record varies by region, and perishable tools are largely absent from it. “The first division” is best understood as a threshold crossed in degrees.
Age divided the work before any person mastered it
Every human economy is organised by age because human skill takes time to build. We have an unusually long childhood, a long period in which bodies grow while knowledge accumulates. Difficult foraging skills do not mature together. Strength, judgement, memory, patience and local knowledge follow different curves.
Children contributed while they learned. Research among Hadza and BaYaka foragers shows children learning subsistence skills in playgroups, teaching and correcting one another as well as learning from adults. In that close-observation study, teaching between children was more frequent than teaching from adults to children. Hadza children gather food from an early age, while their returns and the tasks they attempt change as bodies and skill develop. A systematic archaeological review of hunter-gatherer childhood has identified children's tools, practice pieces, footprints, marks and flintknapping across all inhabited continents. Much of what older archaeology classified as poor workmanship may be evidence of learning.
The same life-course logic continues into adulthood. Tracking and pursuing mobile prey can take years to master. So can reading weather, locating underground foods, knowing when hundreds of plant species are useful, treating illness or producing a reliable tool from difficult stone. Competence in these tasks peaks at different ages. Older adults may contribute less through speed or strength and more through ecological memory, craft, coordination, teaching and care. An economy can therefore be age-structured without assigning anyone a permanent office. The group draws different kinds of value from a person across a life.
Childcare belongs at the centre of this account. Human infants require unusually intensive and prolonged support. Mothers have rarely carried that work alone. Among the Agta of the Philippines, detailed proximity data show care distributed among kin and non-kin; comparative research across small-scale societies likewise finds substantial care by fathers, siblings, grandparents and other camp members. This alloparenting releases time for food production and spreads the cost of raising a slow-growing child. It is a division of reproductive labour as consequential as the division of food-getting. Without it, the complementary foraging system described in the next section is much harder to sustain.
The first lesson is already broader than “he hunted, she gathered”. Work was divided among learners and experts, mobile adults and caregivers, the strong and the knowledgeable. Age created a sequence of different contributions.
Skill made the group larger than any one mind
Specialisation pays when practice matters. Time spent learning one difficult activity can raise returns from that activity, while constant switching keeps everyone mediocre at many things. This is the elementary logic later economists would formalise as comparative advantage, but no calculation is required for a group to discover it. The best tracker goes farther. The most reliable knapper prepares a difficult core. Someone with exact knowledge of a water source leads the move. A person who has delivered many babies is called when labour begins.
Skill also changes what a society can know. No individual has to carry the full technical and ecological repertoire of the group. Knowledge can be distributed among people and recovered through cooperation. This makes a community's effective memory larger than any member's. It also makes the community vulnerable. If only a few people hold a difficult practice, their death or departure can remove it. Models of cultural specialisation capture both sides of the bargain: dividing knowledge can support a more complex collective toolkit, while concentrating knowledge in a few hands increases the risk of loss.
Foragers were nevertheless rarely specialists in the modern occupational sense. Breadth was insurance. A person who could hunt but not find water, repair equipment or recognise edible plants would be dangerously dependent on one uncertain return. Even highly skilled hunters collected fruit, honey and small game. Expertise sat inside general competence.
Skill allocation remained flexible. Many mobile foraging societies have weak formal authority, and leadership often shifts with the task. The person followed on a hunt need not decide where camp moves or how food is distributed. A reputation for competence can create influence without producing an office. This distinction between authority in a practice and rank over persons will matter throughout the history of work.
The division could also cross the edge of a camp. Upper Palaeolithic evidence includes stone and other materials transported over long distances, alongside technologies and styles shared across wide regions. Some objects moved because people moved; others may have passed through visits, gifts or exchange. The remains do not reveal a market or a specialist trader. They do show that a group's practical world could extend beyond the people sleeping at one hearth. Inter-group relationships spread materials, knowledge and access to different ecologies, making the wider network another source of complementary work.
The gender story we told ourselves: “Man the Hunter”
“Man the Hunter” is a gender story built from observations about sex. It took a recurring pattern, with males doing more of some forms of hunting in many documented societies, and turned it into social identities: man as hunter, provider and inventor; woman as gatherer, caregiver and dependent. The heading names that cultural interpretation, not a biological category. This article does not deny sex-linked patterns. It questions the leap from an average pattern to a universal rule about what men and women are, what they may do and whose work drives history.
In April 1966, anthropologists gathered at the University of Chicago for a symposium whose 1968 proceedings, Man the Hunter, carried a title that became a theory of human origins. Hunting, in the simplified public version, was the engine of human evolution. Men ranged, cooperated, made tools and brought meat home. Women remained near camp, constrained by pregnancy and children, gathering dependable foods and receiving the products of male ingenuity.
There is a genuine empirical pattern beneath the caricature. Across many documented foraging societies, men have spent more time pursuing large, mobile game, while women have more often gathered plants and small animals or used methods compatible with infant care. These activities differ in risk, timing, mobility and the age at which peak skill is reached. Complementary foods can improve the household or group's diet; different return profiles can steady consumption; practice can increase efficiency. Reproductive constraints and average differences in strength may affect the costs of some tasks. Cultural rules, prestige and learned expectations affect them too.
The mistake was to turn a tendency into a timeless binary, and then to treat the prestigious half as the whole economy. Gathering, processing, tool maintenance and care leave a fainter archaeological signature than projectile points and butchered bone. Western researchers arrived with their own assumptions about male providers and female domesticity. What the archive preserved and what observers expected reinforced each other.
Recent evidence has made the old certainty untenable. A young woman buried about 9,000 years ago at Wilamaya Patjxa in the Peruvian Andes was interred with a big-game hunting toolkit; a wider review of early American burials found other females associated with hunting tools. Sarah Lacy and Cara Ocobock have assembled archaeological and physiological evidence against treating female hunting as anomalous. An ethnographic review led by Abigail Anderson reported women hunting in most of the 63 societies in its selected sample.
That last result became a lesson in caution of its own. Critics led by Vivek Venkataraman argued that the sample and coding could not establish the headline claim that women hunt in 79 per cent of all foraging societies; the original paper was corrected, and the authors and critics continue to disagree about what the records warrant. Burial goods also require care: a toolkit can mark identity, relationship or ritual rather than the daily activity of the buried person. The strongest conclusion is more precise: women have hunted in many times and places, sometimes deliberately and with specialised tools, and the prevalence, scale and organisation of that hunting varied.
Variation is the substance of the story. Agta women in the Philippines have hunted in groups and with dogs. Martu women in Australia hunt small game in ways that combine food production, land knowledge and the teaching of children. Hadza movement data show strongly patterned differences between men and women, yet men also gather and women also pursue animal foods. Pregnancy, breastfeeding, childcare arrangements, technology, prey, season, local norms and individual skill alter the costs. Sex-linked constraints and gendered expectations both mattered, but neither operated alone or dictated one universal arrangement.
Why divide labour at all?
The first division endured because it solved several problems at once.
It increased proficiency. Time invested in a difficult practice improves performance, especially when the task rewards accumulated knowledge or technique. It allowed parallel work. A camp can gather, hunt, process food, repair equipment and care for children at the same time. It joined complementary returns. Reliable foods and high-variance foods behave differently; pooled together, they can provide both a base and an occasional surplus. It supported dependent people. Children, the ill, the injured and older adults could eat when they did not produce enough that day. And it enlarged collective knowledge. The group could maintain more skills than any individual could learn well.
These benefits depend on transfer. If everyone consumes only what they personally obtain, narrow specialisation is dangerous. Sharing allows one person's return to support another person's task. A formal model of resource sharing shows how sharing alone can generate task specialisation among initially similar agents. Human arrangements are richer than the model, but its logic is useful: the division of labour and the movement of its products make each other viable.
Sharing also manages risk, although risk pooling explains only part of it. Large packages of meat may be difficult for a producer to defend or consume before they spoil. Kin provisioning matters. Reciprocal help can be repaid over long intervals, especially when injury prevents work. Reputation and alliance matter. In some societies, people use levelling norms or assertive demands to prevent successful producers from keeping everything. Brian Wood and Frank Marlowe's Hadza data, for example, found that successful hunters channelled important benefits towards wives, children and kin, challenging a pure “show-off” account of hunting.
Sharing around the fire could be warm, obligatory, strategic, contested and enforced at once. Chimpanzees and other animals share some foods and cooperate in some tasks. What became unusual in our lineage was the scale and flexibility of the package: diverse tasks, long learning, routine transfers, extensive care and cooperation beyond a single mother–offspring pair.
This changes the meaning of the first division. It combined productivity with mutual dependence. The person concentrating on one task could do so because other people covered the tasks left undone, and because some route existed for the results to circulate.
Specialisation · Shared life
A camp works as a network—not a two-person binary.
Specialisation becomes possible because food, care, knowledge and materials circulate across people of different ages, skills and circumstances.
Shared centre
The camp
Mutual dependence makes repeated specialisation sustainable.
A shared camp sits at the centre of a network connecting food acquisition, processing, toolmaking, fire, care, children learning, ecological knowledge, sharing and teaching. Four flows—food, care, knowledge and materials—link the activities.
Cooperation and inequality
Mobile foragers are often described as egalitarian, and many have institutions that resist durable dominance: mobility limits accumulation; food is shared; boastful hunters are mocked; leadership is temporary; people can leave. These are political achievements, not the absence of politics.
These arrangements varied. Along the Pacific coast of North America, abundant, concentrated and storable resources supported sedentary foraging societies with ownership, rank and, in some cases, slavery. A comparative study of 89 societies in that region found institutional inequality associated especially with controllable, clumped aquatic resources. Comparative evidence also associates food storage with religious specialists among hunter-gatherers. Specialisation and hierarchy could both precede farming.
Status could arise without wealth. Expertise attracts attention, trust and influence. A skilled producer may gain allies or desirable partners; a healer may command deference; an elder may control knowledge. Prestige becomes domination or hereditary class only when resources, knowledge and positions can be controlled and transmitted.
This is the point at which the threads of specialisation, status and access first cross. Once knowledge is unequally distributed, access to learning matters. Who is invited to watch? Who has the time, tools and permission to practise? When a task is coded as belonging to one gender, age group or lineage, a practical arrangement can become a gate. The earliest divisions were flexible, but every repeated division carries the possibility of becoming a rule about who may do what.
There was no single first specialist
The phrase “first division of labour” invites a founding scene: one day everyone did everything; the next, some hunted while others gathered. The evidence supports no such boundary.
Task differentiation exists in other social species. Hominins made tools long before Homo sapiens. Provisioning, teaching, care and broad food sharing probably developed at different rates. By the Upper Palaeolithic, the archaeological record shows diverse technologies, long-distance material movement and skill differences that are hard to explain without distributed work and knowledge. Yet even then, most people probably remained capable generalists who moved among many tasks.
The division of labour is therefore less like an invention than a ratchet. Sharing made some differentiation safer. Differentiation rewarded learning. Learning increased dependence on other people's knowledge. Care enabled a longer childhood in which difficult skills could be acquired. Each turn allowed the group to support a little more diversity of practice.
Farming changed the scale and durability of this ratchet. Storage and denser settlement could support people who spent more time producing for others, institutions that controlled redistribution, and roles that persisted beyond a season or individual. The next chapter traces the transformation of flexible task specialisation into more continuous craft, administrative, military and religious work.
What this opening teaches us
That the first division of labour was plural. Age, skill, care, ecology, circumstance and gender all organised work, while reproductive biology shaped some constraints. Any origin story with only a hunter and a gatherer has left most of the camp out.
That specialisation came in degrees. Differentiated tasks became repeated roles, and some repeated roles developed into occupations. The long history of work is partly the history of those arrangements becoming more durable, named and governed.
That learning was itself part of the economy. A long childhood, peer learning, practice beside experts and the memory of older people allowed difficult skills to survive. Human capital existed long before anyone called it capital.
That sharing made specialisation possible. A person can concentrate on one uncertain or time-consuming task only if other people's work reaches them. The division of labour begins in interdependence.
That flexibility was a strength. Foraging roles bent with season, technology, care, injury and opportunity. Variation allowed the system to adapt.
That division can become distinction. Skill brings prestige, and repeated roles can become gates. Hierarchy emerges when resources, authority or access can be held, defended and passed on.
At the old hearth, the expert lifts a prepared core and strikes. A long blade falls cleanly. Nearby, a learner tries the same movement and fails, leaving a hinge in the stone. Someone will fetch fuel before the light goes. Someone is carrying an infant. Someone knows where the herd moved after rain. No one does everything, and no single division explains the whole.
The oldest durable fact about work is human dependence on differences of age, knowledge, strength, attention and circumstance. A career is a distant descendant of that arrangement. Long before work acquired titles, wages or institutions, it was already a promise between specialists: I can learn this because you are learning that.
Notes on the evidence
There is no direct evidence for a first moment at which human labour was divided, and this article does not assign one. Archaeology preserves skill and activity unevenly; ethnography documents recent societies with their own histories. The Upper Palaeolithic division-of-labour hypothesis is associated with Steven Kuhn and Mary Stiner (2006) and remains debated. Evidence for age-structured learning comes from archaeological reviews and studies among Hadza and BaYaka children; evidence for distributed childcare comes from Agta and comparative alloparenting research. The sexual division of foraging labour is both patterned and variable. Anderson and colleagues' 2023 ethnographic review, the criticism by Venkataraman and colleagues, and archaeological arguments by Haas, Lacy and Ocobock should be read together. Explanations of male hunting as provisioning, risk pooling or signalling remain live alternatives rather than a settled choice. Claims about forager egalitarianism are qualified by evidence for hierarchy where valuable resources could be stored and controlled.
Sources and further reading
- Sileshi Semaw et al., “2.5-Million-Year-Old Stone Tools from Gona, Ethiopia,” Nature 385 (1997), 333–336. Article
- Steven L. Kuhn & Mary C. Stiner, “What's a Mother to Do? The Division of Labor among Neandertals and Modern Humans in Eurasia,” Current Anthropology 47 (2006), 953–980. Article
- Sheina Lew-Levy et al., “Who Teaches Children to Forage? Exploring the Primacy of Child-to-Child Teaching among Hadza and BaYaka Hunter-Gatherers,” Evolution and Human Behavior 41 (2020), 12–22. Article
- Annemieke Milks, Sheina Lew-Levy, Noa Lavi, David E. Friesem & Rachel Reckin, “Hunter-Gatherer Children in the Past: An Archaeological Review,” Journal of Anthropological Archaeology 64 (2021), 101369. Article
- Alyssa Crittenden et al., “Juvenile Foraging among the Hadza,” Evolution and Human Behavior 34 (2013), 299–304. Article
- Abigail E. Page et al., “Testing Adaptive Hypotheses of Alloparenting in Agta Foragers,” Nature Human Behaviour 3 (2019), 1154–1163. Article
- Karen L. Kramer & Amanda Veile, “Infant Allocare in Traditional Societies,” Physiology & Behavior 193 (2018), 117–126. Record
- Richard B. Lee & Irven DeVore (eds.), Man the Hunter (Aldine, 1968). Record
- Abigail Anderson et al., “The Myth of Man the Hunter: Women's Contribution to the Hunt across Ethnographic Contexts,” PLOS ONE 18 (2023), e0287101, with correction and subsequent debate. Article
- Vivek V. Venkataraman et al., “Woman the Hunter?,” Evolution and Human Behavior (2024), a methodological critique of the 2023 ethnographic review. Record
- Randy Haas et al., “Female Hunters of the Early Americas,” Science Advances 6 (2020), eabd0310. Article
- Sarah Lacy & Cara Ocobock, “Woman the Hunter: The Archaeological Evidence,” American Anthropologist 126 (2024), 19–31. Article
- Brian M. Wood et al., “Gendered Movement Ecology and Landscape Use in Hadza Hunter-Gatherers,” Nature Human Behaviour 5 (2021), 436–446. Article
- Michael Gurven & Kim Hill, “Why Do Men Hunt? A Reevaluation of ‘Man the Hunter’ and the Sexual Division of Labor,” Current Anthropology 50 (2009), 51–74. Article
- Brian M. Wood & Frank W. Marlowe, “Household and Kin Provisioning by Hadza Men,” Human Nature 24 (2013), 280–317. Record
- Jan J. Kreider et al., “Resource Sharing Is Sufficient for the Emergence of Division of Labour,” Nature Communications 13 (2022), 7232. Article
- Eric Alden Smith & Brian F. Codding, “Ecological Variation and Institutionalized Inequality in Hunter-Gatherer Societies,” PNAS 118 (2021), e2016134118. Article
- Joseph Watts et al., “Food Storage Facilitates Professional Religious Specialization in Hunter-Gatherer Societies,” Evolutionary Human Sciences 4 (2022), e17. Article
- Yotam Ben-Oren, Oren Kolodny & Nicole Creanza, “Cultural Specialization as a Double-Edged Sword,” Philosophical Transactions of the Royal Society B 378 (2023), 20210418. Article
- Ofer Bar-Yosef, “The Archaeological Framework of the Upper Paleolithic Revolution,” Diogenes 54 (2007), 3–18. Article